I am trying to implement FastRCNN using Mindspore, I am using the code from the offciel gitee as a reference: 'https://gitee.com/mindspore/mindspore/tree/r1.0/model_zoo/official/cv/faster_rcnn', everything in the code is working well but when I start training it gives me the follwing error:
raise ValueError(f"For '{prim_name}', the x_shape {x_shape} and y_shape {y_shape} can not broadcast.")
ValueError: For 'Add', the x_shape [2, 256, 24, 32] and y_shape [2, 256, 48, 80] can not broadcast.
Knowing that I am using VOC dataset and MindSpore 1.3.0
This a snapshot of the code:
## The shape of the mindrecord:
fasterrcnn_json = {
"image": {"type": "bytes"},
"annotation": {"type": "int32", "shape": [-1, 6]},
##### annotation=[x1,y1,x2,y2,label(int),iscrowd(int)]
}
writer.add_schema(fasterrcnn_json, "fasterrcnn_json")
###############the function of creating the dataset:
def create_fasterrcnn_dataset(mindrecord_file, batch_size=2, repeat_num=12, device_num=1, rank_id=0,
is_training=True, num_parallel_workers=4):
"""Creatr FasterRcnn dataset with MindDataset."""
ds = de.MindDataset(mindrecord_file, columns_list=["image", "annotation"], num_shards=device_num, shard_id=rank_id,
num_parallel_workers=1, shuffle=is_training)
decode = C.Decode()
ds = ds.map(operations=decode, input_columns=["image"], num_parallel_workers=1)
compose_map_func = (lambda image, annotation: preprocess_fn(image, annotation, is_training))
hwc_to_chw = C.HWC2CHW()
normalize_op = C.Normalize((123.675, 116.28, 103.53), (58.395, 57.12, 57.375))
horizontally_op = C.RandomHorizontalFlip(1)
type_cast0 = CC.TypeCast(mstype.float32)
type_cast1 = CC.TypeCast(mstype.float16)
type_cast2 = CC.TypeCast(mstype.int32)
type_cast3 = CC.TypeCast(mstype.bool_)
if is_training:
ds = ds.map(operations=compose_map_func, input_columns=["image", "annotation"],
output_columns=["image", "image_shape", "box", "label", "valid_num"],
column_order=["image", "image_shape", "box", "label", "valid_num"],
num_parallel_workers=num_parallel_workers)
flip = (np.random.rand() < config.flip_ratio)
if flip:
ds = ds.map(operations=[normalize_op, type_cast0, horizontally_op], input_columns=["image"],
num_parallel_workers=1)
ds = ds.map(operations=flipped_generation,
input_columns=["image", "image_shape", "box", "label", "valid_num"],
num_parallel_workers=num_parallel_workers)
else:
ds = ds.map(operations=[normalize_op, type_cast0], input_columns=["image"],
num_parallel_workers=1)
ds = ds.map(operations=[hwc_to_chw, type_cast1], input_columns=["image"],
num_parallel_workers=1)
else:
ds = ds.map(operations=compose_map_func,
input_columns=["image", "annotation"],
output_columns=["image", "image_shape", "box", "label", "valid_num"], ###what is valid-num
column_order=["image", "image_shape", "box", "label", "valid_num"],
num_parallel_workers=num_parallel_workers)
ds = ds.map(operations=[normalize_op, hwc_to_chw, type_cast1], input_columns=["image"],
num_parallel_workers=1)
# transpose_column from python to c
ds = ds.map(operations=[type_cast1], input_columns=["image_shape"])
ds = ds.map(operations=[type_cast1], input_columns=["box"])
ds = ds.map(operations=[type_cast2], input_columns=["label"])
ds = ds.map(operations=[type_cast3], input_columns=["valid_num"])
ds = ds.batch(batch_size, drop_remainder=True)
ds = ds.repeat(repeat_num)
ds
return ds
######Dataset creation:
dataset = create_fasterrcnn_dataset(mindrecord_file, repeat_num=1,
batch_size=config.batch_size, device_num=device_num, rank_id=rank)
##### Start trainign the model:
model.train(config.epoch_size, dataset, callbacks=cb)-How do you think I can solve the error?
-The question also is, How the model suppose to split the data into X and Y and what is why in this case?
Thank you in Advance!
I attached the full error traceback as doc file.
